An intelligent, RAG-based (Retrieval-Augmented Generation) HR Helpdesk application designed to answer employee queries about company policies instantly and accurately.
This solution was architected and implemented using Antigravity, Google's advanced agentic AI coding assistant. Leveraging autonomous capabilities, Antigravity handled the end-to-end development lifecycle:
- Rapid Prototyping: Instantly scaffolded a production-ready Full Stack (Next.js + FastAPI) architecture.
- Complex Problem Solving: Autonomously diagnosed and resolved system-level issues, including Python environment conflicts and IPv6 networking constraints.
- Agentic Verification: Deployed autonomous browser subagents to verify user flows, record demos, and ensure a bug-free experience before deployment.
Experience the seamless interaction flow:
- Upload: Admins upload PDF policy documents via the Admin Console.
- Ask: Employees ask natural language questions in the Chat Interface.
- Answer: The AI retrieves relevant policy sections and provides precise answers.
- 🤖 AI-Powered Chat: Uses Gemini 2.5 Flash to answer questions based only on your verified HR documents (RAG).
- 📂 Admin Console: Easy-to-use interface for HR administrators to upload and manage policy PDFs.
- 🧠 vector Search: Powered by ChromaDB for semantic search and retrieval of relevant document chunks.
- ⚡ Modern Stack: Built with FastAPI (Python) for high-performance backend and Next.js (TypeScript) for a responsive frontend.
- 🔌 Extensible: Designed with a modular architecture to easily swap LLMs or vector stores.
- Frontend: Next.js 16, Tailwind CSS, Lucide React
- Backend: FastAPI, Uvicorn, Python 3.11
- AI & ML: Google Gemini API, LangChain, SentenceTransformers
- Database: ChromaDB (Vector Store)
- Ingestion: Admin uploads PDFs -> Text extracted -> Split into chunks -> Embeddings generated (all-MiniLM-L6-v2) -> Stored in ChromaDB.
- Retrieval: User asks question -> Query embedding -> Similarity search in ChromaDB -> Top k chunks retrieved.
- Generation: Retrieved chunks + User Query -> Gemini LLM -> Precise Answer.
- Python 3.11+
- Node.js 18+
- Google Gemini API Key
cd backend
# Create virtual environment
python -m venv venv
# Windows:
.\venv\Scripts\activate
# Mac/Linux:
# source venv/bin/activate
# Install dependencies
pip install -r requirements.txt
# Configure Environment
cp .env.example .env
# Open .env and add your GOOGLE_API_KEYcd frontend
# Install dependencies
npm install
# Configure Environment
cp .env.local.example .env.localStart Backend:
cd backend
python -m uvicorn main:app --reload --host 0.0.0.0 --port 8000Backend runs at: http://localhost:8000
Start Frontend:
cd frontend
npm run devFrontend runs at: http://localhost:3000
MIT